Tectonic and Seismic Implications of an Intersegment Rupture the Damaging May 11Th 2011 Mw 5.2 Lorca, Spain, Earthquake

Tectonophysics 546-547 (2012) 28–37

Contents lists available at SciVerse ScienceDirect

Tectonophysics

journal homepage: www.elsevier.com/locate/tecto

Tectonic and seismic implications of an intersegment rupture The damaging May 11th 2011 Mw 5.2 Lorca, Spain, earthquake

José J. Martínez-Díaz a,⁎, Marta Bejar-Pizarro b, José A. Álvarez-Gómez a,c, Flor de Lis Mancilla d,e, Daniel Stich d,e, Gerardo Herrera b, Jose Morales d,e a Dpto. de Geodinamica, Universidad Complutense, IGEO (UCM-CSIC), Calle Jose A, Novais 2, 28040 Madrid, Spain b InSARlab Geohazards InSAR laboratory Geohazards group Geoscience Research dept., Geological Survey of Spain C/Alenza 1 28003 Spain c Instituto de Hidráulica Ambiental “IH Cantabria”, Universidad de Cantabria, E.T.S.I. Caminos, Canales y Puertos, Santander, Spain d Instituto Andaluz de Geofísica, Campus Universitario de Cartuja, Universidad de Granada, Granada, Spain e Departamento de Fisica Teorica y del Cosmos, Universidad de Granada, Granada, Spain article info abstract

Article history: On May 11th 2011, a Mw 5.2 earthquake stroke the city of Lorca in the SE Spain. This event caused 9 fatalities, Received 9 February 2012 300 injuries and serious damage on the city and the surrounding areas. The Lorca earthquake occurred in the Received in revised form 5 April 2012 vicinity of a region bounding two well-known segments of a large active fault, the Alhama de Murcia fault Accepted 11 April 2012 (AMF). The Lorca earthquake offers a unique opportunity to study how strain is accommodated in an inter- Available online 24 April 2012 segment region of a large strike slip fault. We map recent tectonic structures in the epicentral region and we use radar interferometry to analyze the coseismic deformation. Combining these data with seismological ob- Keywords: ﬁ ﬂ Betic Cordillera servations of Lorca seismic sequence we rst model the source of the earthquake. Then we analyze the in u- INSAR ence of our preferred model in the adjacent segments by Coulomb failure stress modeling. The proposed Coulomb stress transfer earthquake source model suggests that this event ruptured an area of ~4×3 km within the complex structure Seismic hazard that limits the Goñar–Lorca and Lorca–Totana segments of the AMF. The induced static stress change on the Active tectonics adjacent segments of the fault represents a seismic cycle advance equivalent to 200 to 1000 years of tectonic Intersegment zone loading. © 2012 Elsevier B.V. All rights reserved.

1. Introduction zones behave as relaxation barriers and large ruptures skip over the intersegment area that remains unbroken (Das and Aki, 1977; Active faults are organized in more or less uniform segments sep- Scholz, 1990). The study of small earthquakes that break these in- arated by intersegment regions, characterized either by a change in tersegment areas gives us the opportunity to analyze this complex the geometry of the fault or by the presence of structural complexities behavior. (Elliott et al., 2012; Fliss et al., 2005; Harris and Day, 1993; King, On May 11th 2011, a Mw 5.2 earthquake stroke the city of Lorca in 1986; Klinger, 2010; Shengji et al., 2011; Wesnousky, 2006). Segmen- south-eastern Spain. This earthquake occurred 2 h after a Mw 4.6 tation is important because of its implications for the rupture behav- foreshock and caused 9 fatalities, 300 injuries, serious damage on ior during earthquakes. Segments are fault slip prone areas during 1164 buildings and economic losses over 1200 M€ (data from the large earthquakes (DePolo et al., 1989; 1991), whereas intersegment Municipality of Lorca updated November 2011). The Lorca earth- zones are deﬁned as areas where rupture begins or stops during an quake is especially signiﬁcant because it occurred in the vicinity of a earthquake (e.g. Aochi et al., 2000; Jackson et al., 2006; Klinger region bounding two well-known segments of a large active fault, et al., 2005; Lozos et al., 2011). This complex behavior is observed the Alhama de Murcia fault (AMF) (Fig. 1). This fault is the source not only in the horizontal rupture propagation, but also in the rupture of Mw>6.5 historical and pre-historical earthquakes (Martinez-Diaz propagation at depth (Elliott et al., 2011; Jackson et al., 2006; Li et al., et al., 2001). The AMF accommodates ~0.1–0.6 mm/year of the 2011; Nissen et al., 2010). During major and less frequent earth- approximately 5 mm/year of convergence between African and quakes, several segments can slip at a time, whereas the intersegment Eurasian plates (Masana et al., 2004) and belongs to the Eastern zones can behave as areas of high slip release. This occurred during Betics Shear Zone (Silva et al., 1993). The AMF is one of the largest the Wenchuan earthquake (Shen et al., 2009). In other cases these faults of this shear zone. Most of the largest damaging historical earthquakes are related with this structure (Fig. 1). This fault presents aNE–SW direction; it is ~100 km length and is divided into 4 seg- ⁎ Corresponding author at: Calle Jose A, Novais 2, 28040 Madrid, Spain. Tel.: +34 – – – 913944835; fax: +34 913944634. ments: Goñar Lorca; Lorca Totana; Totana Alhama de Murcia and E-mail address: [email protected] (J.J. Martínez-Díaz). Alhama de Murcia–Alcantarilla (Fig. 1B). Paleoseismic studies along

4° W 2° W 0°

Iberian peninsula M editerranean sea

39°N cordillera Betic Tell Rif African plate

BSF

38°N

Intensity (EMS) VI

PF 37°N VII VIII AMF IX CF X

36°N A

050 100 200 Kilometers

M Murcia Alcantarilla IAG IGN HAR Fortuna basin

EXAH A-A

N F Alhama Espuña Range EMUR IAG de Murcia Carrascoy Range EXZA 37°50' VALD

Totana Lorca basin T- A

S T

Gualdalentín valley Lorca L-T

Las Estancias UMV6 N Range

37°35’ 37°35’ VELZ Pto. Lumbreras MAZA G-L S E Quaternary deposits

l Neogene deposits B 0 10 20 Km

2° W 1° 45'W 1° 30'W 1° 15'W

Fig. 1. A) Location map of the study area in which the Quaternary active faults are projected. Circles represent the historical seismicity with intensity (EMS)>VI (data from the Instituto Geografico Nacional). The ellipse indicates the position of the Eastern Betic Shear Zone, CF: Carboneras fault; PF: Palomares fault; AMF: Alhama de Murcia fault; CCF: Carrascoy fault; BSF: Bajo Segura fault. B) Map of the Alhama de Murcia fault, arrows indicates the limits of the four main segments of this fault: GL: Goñar–Lorca segment, LT: Lorca–Totana Segment, TA: Totana–Alhama segment; AA: Alhama–Alcantarilla segment. ST: Sierra de La Tercia. The star and the circles are the mainshock and aftershocks of the Lorca 2011 seismic sequence taken from Lopez-Comino et al. (2012). Focal solutions of the foreshock (F) and the mainshock (M) from several agencies are shown, IAG: Instituto Andaluz de Geofisica; IGN: Instituto Geografico Nacional; HARV: Harvard University. In both maps the black triangles are the seismic stations utilized in the aftershock relocation.

this fault suggest that the two segments converging on Lorca (Goñar– The Lorca earthquake offers a unique opportunity to study how Lorca and Lorca–Totana) ruptured during the Quaternary as a single a strain is accommodated in this region that includes not only the seismogenic source producing earthquakes of Mw 6.9–7.3 (Masana AMF but also secondary active structures. We map recent tectonic et al., 2005; Ortuño et al., 2012). Other paleoearthquakes identified structures in the epicentral region and we use radar interferom- on these segments are smaller (Mw~6) and seem to have ruptured etry to constrain the coseismic deformation. Combining these data only one segment (Masana et al., 2004). The Lorca intersegment with published seismological data of the seismic sequence of Lorca area could play a significant role in the seismogenic behavior of the (Lopez-Comino et al., 2012) we model the earthquake source. Using AMF. Coulomb Failure Stress transfer (ΔCFS) models we analyze the 30 J.J. Martínez-Díaz et al. / Tectonophysics 546-547 (2012) 28–37 influence of our preferred source in the adjacent segments. We finally the other nodal plane perpendicular to the AMF dipping to the SW compare our results with topography and fault structure close to the (IAG, 2011; IGN, 2011)(Fig. 2). The latter is difficult to explain from rupture area and discuss the implications for strain accommodation, the local structure. The former, on the other hand, is parallel to sev- fault behavior and seismic hazards in the region. eral branches of the AMF in the intersegment zone. Aftershocks regis- tered until the 7th July were relocated by Lopez-Comino et al. (2012) using a dense seismic station network (Fig. 2). The aftershock epi- 2. Structure of the epicentral area centers aligned parallel to the AMF and concentrated on the north of the intersegment zone. These evidences suggest that the source The Lorca 2011 earthquake occurred near the intersegment of the earthquake is parallel to the AMF dipping to the north as pro- zone located between Goñar–Lorca and Lorca–Totana segments posed in a preliminary study by Vissers and Meijninger (2011).In (Fig. 1). A field survey conducted in the epicentral area 2 days this work we use InSAR measurements of the coseismic deformation after the earthquake concluded that the Lorca earthquake did not to better define the earthquake source parameters. rupture the surface (IGME, 2011).Weperformeddetailedmapping of recent structures on the epicentral area to understand the kine- matic of structures in the intersegment zone and to assess the 3. Insar analysis potential source for the Lorca earthquake. For this purpose we use field data, aerial photography and a digital elevation model Immediately after the occurrence of the earthquake Frontera et al. derived from LIDAR (Fig. 2). In this area the AMF undergoes a (2012) made a DInSAR measurement of the coseismic deformation change of direction from N 55° to the northeast, to N 35° to the using a pair of TerraSAR-X images, and a theoretical simple numerical southwest. The structure of the fault is rather complex in this re- model based on estimated seismic rupture dislocation. They found gion, with a branched geometry due to the existence of contrac- 3 cm of vertical deformation in the northern wall of the AMF. In this tional strike-slip duplex structures (Martinez-Diaz, 2002)andthe chapter we present a complete InSAR analysis considering uniform interaction with the Las Viñas Fault to the north of Lorca. This is and distribute slip modeling in order to understand the coseismic de- a WNW secondary fault that connects with the AMF generating a formation induced by the earthquake in the frame of the local tecton- contracting slice raised by the movement of the AMF, which causes ic structure of the Alhama de Murcia Fault. the lifting of the NE corner of the Sierra de Las Estancias (Fig. 2). We use 5 Envisat ASAR images from one descending track to form Another example is the interaction between contemporary re- 4 coseismic interferograms. The image acquisition times and inter- verse, normal and strike-slip minor faults, and the disruption of ferograms constructed for this study are shown in Table 1. We can the SW termination of the Sierra de La Tercia anticline. recognize the coseismic signal in each interferogram, but the phase The converging segments of the AMF in Lorca zone (GL and LT in difference also contains residual orbital errors and atmospheric Fig. 2) present a much simpler trace in the field and are observed dip- phase delays (Fig. 3). To reduce these artifacts and improve the ping 55°–75° NW, and bounding the mountain fronts of the ranges. signal-to-noise ratio, we correct each interferogram of an orbital The reverse component of the movement produced this topography ramp and a phase/elevation correlation. Then we calculate an average since upper Miocene (Martinez-Diaz, 2002). However, there is no re- interferogram using the 4 corrected interferograms (e.g. Cavalie et al., lief in the intersegment zone where a depressed region dominates the 2007). See Figs. SM1 and SM2 of the supplementary material for morphology (Fig. 2). details. Published focal mechanism for the Lorca 2011 earthquake present The resulting average interferogram is shown in Fig. 4. The ob- a nodal plane sub parallel to the AMF dipping to the north, being served displacement along the line of sight (LOS) direction (white

D SW A-A' NE M F Sierra de La Tercia 1000 m Sierra de Las Estancias IAG 500 m IAG IGN HAR A' 5 10 15 Km

37.73°N B-B' NW SE Lorca Basin Sierra de La Tercia 1000 m Sierra de Las Estancias C AMF 500 m 2 4 6 Km

NW C-C' SE 37.7°N D' 1000 m Lorca basin AMF 500 m B LasV iñas Fault 5 10 15 Km River Guadalentín Lorca

D-D' Sierra de NW SE LORCA

1000 m 37.67°N Sierra de La Tercia AMF Las Estancias 500 m C' A 2 4 6 Km B' 5 Km

1.75° W 1.7° W 1.65° W

Fig. 2. Map of the detailed structure of the Alhama de Murcia fault in the epicentral area of the Lorca 2011 earthquake. Foreshock (F) mainshock (M) and relocated aftershocks from Lopez-Comino et al. (2012) are projected, together with the available focal solutions (see caption of Fig. 1). Mapped structures affecting the upper Miocene and Pliocene deposits are also shown: dotted lines: fold axis; continuous lines: normal faults; lines with arrows: reverse-strike slip faults. AMF (G–L): Goñar-Lorca segment; AMF (L–T): Lorca–Totana segment. To the left four transversal and longitudinal topographic proﬁles are shown. J.J. Martínez-Díaz et al. / Tectonophysics 546-547 (2012) 28–37 31

Table 1 arrow in Fig. 4) shows a region of ~5 km diameter to the NW of the Interferograms constructed for this study. AMF that moves towards the satellite (maximum of ~2 cm) and a sig- Interferograms Date1 Date2 (*)Bperp niﬁcant larger region to the SE of the AMF that moves away from the satellite. The kind and size of the ground deformation to the north of Int1 27-Nov-10 26-May-11 70 Int2 26-Apr-11 26-May-11 −100 the AMF are consistent with the coseismic deformation expected for Int3 26-Apr-11 25-Jun-11 80 the Mw 5.2 Lorca earthquake. Besides, roughly the same deformation Int4 26-Apr-11 25-Jul-11 17 is present in two independent interferograms (Fig. 3), suggesting that (*)Bperp: perpendicular baseline in meters. it is related to the same phenomena. On the other hand, the

110426_110526 110426_110625 37.75o A A

B' B'

37.7o

B B

A' A'

37.65o

A B -1.75o -1.7o -1.65o -1.75o -1.7o -1.65o

110426_110725 101127_110526 37.75o

A A

B' B'

37.7o

B B

A' A'

37.65o

cm LOS C D "3"2"1 0 1 2 3 4 -1.75o -1.7o -1.65o -1.75o -1.7o -1.65o Distance (km) Distance (km) 0510 0246 4 A A' 4 B B'

2 2

LOS (cm) LOS (cm) 0 0

110426_110526 110426_110625 110426_110725 101127_110526

Fig. 3. Original coseismic interferograms used in this study. Dates of the two images combined in each case are indicated on top of each interferogram (notation yymmdd). Same color scale is used in all of them. Black lines represent tectonic structures mapped in the area. Brown star is the epicenter of the Mw 5.1 Lorca earthquake. Cross sections AA′ and BB′ for each interferogram are also shown. The coseismic signal can be recognized in each interferogram, but the phase difference also contains residual orbital errors and atmospheric phase delays. 32 J.J. Martínez-Díaz et al. / Tectonophysics 546-547 (2012) 28–37

o AB C

o D

o o o o o o

Fig. 4. a) Average interferogram showing surface displacement associated with the Mw 5.2 2011 Lorca earthquake. Individual interferograms were processed using the Caltech/JPL (Pasadena, CA, USA) repeat-orbit interferometry package (ROI PAC) (http://www.roipac.org/). The topographic phase contribution was removed using a 90 m DEM from NASA's SRTM. The orbital information used in the processing was provided by the European Space Agency (DORIS orbits). The color scale refers to change in the radar line-of-sight (LOS) direction. Positive displacements are associated with a range decrease and negative displacements are associated with a range increase (movement towards and away the satellite, respectively). The satellite to ground radar line-of-sight (LOS) is shown with a white arrow, which is inclined ~24° from the vertical. (b) Preferred models of coseismic slip from the Lorca earthquake constrained using InSAR data. Blue rectangle represents the preferred uniform-slip model and color scale shows the preferred slip distribution model. Artifacts near the edges of the slip distribution model caused by poor resolution have been masked (see Fig. DR7 for details). c) and d) NW–SE cross-section perpendicular to the AMF. c) LOS deformation observed (red points) and modeled (blue points). d) Black line represents the fault plane used in the slip distribution inversion and thick blue line represents the fault from uniform-slip model. Red star represents the location of the mainshock and black circles represent aftershocks (Lopez-Comino et al., 2012). Black rectangle limits the region covered by data used in inversion. Red lines represent the AMF trace.

deformation to the SE of the AMF covers a much larger area than constraint on the mechanism of small earthquakes (e.g. Lohman and expected for a Mw 5.2 earthquake, and it is also present in non- Simons, 2005). To reduce the non-linearity we use both seismic data coseismic interferograms. The amplitude of this deformation varies and a priori information on the fault geometry to constrain some of depending of the temporal span of the interferogram, suggesting the parameters. that the phenomenon is not related to the Lorca earthquake. A previ- The strike of the fault plane is fixed at 235°. This value is taken ous work using InSAR temporal series recognized an important from the orientation of the Alhama de Murcia Fault (AMF) at the groundwater-related land subsidence in this region (Gonzalez and latitude of Lorca and is consistent with published values for the Fernández, 2011) during the period 1992–2007, that may be the strike of the focal mechanism parallel to the AMF (230° from cause of the deformation present in our interferograms SE of the Harvard CMT solution, 230° from IGN solution; 240° from IAG so- AMF. Therefore, we assume that the deformation SE of the AMF is lution). We fixed the rake at 39°, which corresponds to a sense of not associated with the Lorca earthquake and thus it was not consid- slip vector oriented 196°, consistent with published values (196° ered to build coseismic models. from IAG solution, 197° from IGN solution, 198° from Harvard Note that all the interferograms include some time after the main- CMT solution). shock (from a minimum of 15 days in the Int1 and Int2 interfero- We explore a series of different values for the position, dip and up- grams, to a maximum of 75 days in the Int4 interferogram, dip limit of the fault plane in 4 consecutive searches using both seis- (Table 1). It is therefore possible that they include some post- mic data and a priori information on the fault geometry to constrain seismic deformation together with the coseismic deformation. some of the parameters (see Fig. 5 for details). When exploring the dip we obtain a wide interval of possible values: 55°–70°, suggesting 3.1. Uniform-slip models that our dataset is not very sensitive to the dip of the fault plane. We choose 55° because it is consistent with dip values for the fault plane To explain the deformation pattern we model the earthquake as a parallel to the AMF from Harvard CMT solution and IAG solution (52° dislocation in an elastic medium (Okada, 1985). We prepare InSAR and 54° respectively). Moreover the 55° dip is consistent with the data for inversion by reducing the number of points to about 1500 location of the mainshock and aftershocks (Fig. 4D). Preferred values using a uniform sampling. Black dashed rectangle in Fig. 4 shows for each parameters and their standard deviation are shown in InSAR data used for the models (a region of ~4.5×6.5 km and a spac- Table 2. The histograms shown in Fig. 5 reveal the estimated uncer- ing of ~145 m). tainties for each explored parameter. The preferred uniform-slip We use an inversion procedure based on a least-square minimiza- fault plane is represented in Fig. 4B and D (blue rectangle and blue tion algorithm developed by Tarantola and Valette (1982) that assumes line, respectively). uniform slip on a rectangular fault plane using nine parameters (strike, Our preferred model fits well with the observed deformation pat- dip, rake, length, bottom and top depth, average slip and geographical tern. However, according to our model, the hypocenter of the main- coordinates of the plane). Determining the nine mutually depen- shock is out of the rupture plane (Fig. 4D). We performed models dent parameters of the fault plane is a highly non-linear process. Be- where spatial slip variations along the rupture plane are allowed to sides, we only have interferograms from one satellite line-of sight check if InSAR data are compatible with some significant slip occur- direction, i.e. a single component of deformation, which provides little ring in the hypocentral area. J.J. Martínez-Díaz et al. / Tectonophysics 546-547 (2012) 28–37 33

37.75o 40%

20%

Search 1

0% 0246 h1 37.7o 100% 100% 80% 80% 60% 60% 40% 40% Search 2 20% 20% 0% 0% 1.7 0 1.68 1.66 37.68 37.72 o lon lat 37.65

cm LO S 20 2

3 40 % 1.75o 1.7o 1.65o 20 %

0% 50 60 70 80 Dip

40 % 40% 40% 40%

20 % 20% 20% 20%

Search 0% 0% 0% 0% 0 5 10 0 2 0 2 4 6 8 10 12 48 Lenght h1 h2 M0*1e+16(N.m)

Fig. 5. Schematic diagram illustrating the exploration of the fault plane geometry. We explore the parameters of the fault plane in 4 consecutive searches using both seismic data and a priori information on the fault geometry to constrain some of the parameters. In each search, parameters fixed, explored and free during the inversion procedure are indicated. On the right, we indicate the output of each exploration. On the left, histograms to depict the distribution of the model parameters fixed or constrained in the explorations are shown The map at the top right shows the explored values of longitude and latitude (Lon and Lat, which refer to the location of the center of the upper part of the fault plane projected diagonally to the surface). During Search1, only Lat and Lon positions located on the mapped fault traces are used in the models (white crosses). In the Search 2, once the h1 is constrained, Lon and Lat are inverted, to find the location of the fault plane that better fits the InSAR observations. Preferred values for Lat and Lon are outside the mapped fault traces (red circles in the map). This might suggest that the rupture plane corresponds to a branch of the fault not completely developed (and thus it does not reach the surface yet). Preferred values for each parameters and their standard deviation are shown in Table 2.

3.2. Distributed-slip models little or no slip during the mainshock. The fit to the InSAR observations does not improve compared to that of the uniform slip model, suggest- We extend the previously determined fault plane along strike ing that the rupture is fairly regular, but the distributed-slip model pro- and down-dip and we divide it into an array of 20×20 elements of vides better compliance with seismic data. The value of Mo according to ~0.5 km by 0.5 km. To solve the slip distribution along these 400 this geodetic model is 5.177E +16 Nm (assuming a shear modulus of patches we use a least-squares minimization with the non-negativity 30 GPa). It is therefore compatible with the seismic moment calculated constraint on the slip, imposing the rake of 39°. To limit oscillations of by Lopez-Comino et al. (2012). the solution, we also impose some smoothing on the solution, by mini- mizing the second-order derivative of the fault slip (e.g.; Du et al., 1992; 4. CFS static stress transfer models Harris and Segall, 1987; Simons et al., 2002). We determine the smoothing factor from a trade-off curve that balances both the model rough- We used the equations of Okada (1992) to obtain the strain field in ness and data misfit(Fig. 6). Fig. 4B shows the best coseismic slip the vicinity of the earthquake rupture. The Okada model computes the distribution model from InSAR data inversion. This is characterized displacement due to a rectangular dislocation on an elastic half-space. by a zone of maximum slip of about 3 km by 3 km, with a maximum We used a young modulus of 8×1010 Pa and a Poisson ratio of 0.25, value of 15 cm. This area roughly coincides with the uniform slip fault which is equivalent to a shear modulus of 3.2×1010 Pa. From this strain plane (blue rectangle in Fig. 4B). The distributed-slip model suggests that ~4 cm of slip could have occurred near the hypocenter of the mainshock. The aftershocks are concentrated in the down dip termination of the rupture, in areas that have

Table 2 Source parameters for the uniform‐slip fault plane. Strike, dip and rake parameters are ﬁxed in the inversions (235°, 55° and 39° respectively). See text for details.

Parameter Value

*Lon −1.6798±0.004 *Lat 37.6884±0.003 Depth to top (km) 1.5±0.5 Depth to bottom (km) 4.9±1.5 Fault length (km) 3±1.7 Moment (N∗m) 4.396E+16±7.0E+15 Fig. 6. L2 norm of least squares inversion misﬁt versus model roughness. Arrow indi- *Lon and Lat are in geographical coordinates and refers to the location of the cates the location of the optimal smoothing parameter where the balance between center of the upper part of the fault plane projected diagonally to the surface. model misﬁt and smoothness is achieved. 34 J.J. Martínez-Díaz et al. / Tectonophysics 546-547 (2012) 28–37

o o o AB C’ B B’

o o

B’ CC’

o o o o o

Fig. 7. Coulomb Failure Stress (CFS) change computed on fault planes NE–SW. A) Map of dCFS at 6 km depth (shown as a dashed line in B and C); the dashed lines in A show the cross sections B–B′ and C–C′ (B and C subfigures respectively). The circles show the aftershocks. The rectangle shows the mainshock rupture main patch. The surface traces of the structures are shown as thin lines. Scale of B) and C) in km. we obtain the Coulomb Failure Stress change (ΔCFS) defined by the Robinson and McGinty, 2000; Stein, 1999). The modification of this pa- equation: rameter influences the amount of stress change due to the normal stress variation as it can be seen from Eq. (1). A variation between 0.3 and 0.6 ΔCFS ¼ Δτ−μ′⋅Δσ ð1Þ does not produce significant changes on the results, except for a slight modification on the off-fault lobes of the stress change. where Δτ and Δσ are the shear and normal stress variations on the The convenience of using a constant apparent friction coefficient fault plane respectively, and μ′ is the apparent friction coefficient is usually discussed and seems reasonable for fault zones that show (Harris, 1998; Reasemberg and Simpson, 1992). This apparent (or effec- a different hydraulic behavior compared to the host rock (Beeler tive) friction coefficient is defined as: et al., 2000), due to the high anisotropy in the damage zone (Cocco and Rice, 2002). In undrained situation during the short term post- μ′ ¼ μ⋅ðÞ1−B ð2Þ seismic response, pore pressure changes are proportional to normal stress changes on the fault. Hence, the pore pressure variation can where μ is the friction coefficient and B is the Skempton coefficient be included in the right term in Eq. (1) by means of Eq. (2). which varies between 0 and 1. This coefficient introduces the role of We calculate static stress changes at 6 km depth on planes with pore pressure on the Coulomb failure function. The values of the ap- the same orientation as the rupture (Fig. 7A–C), which represents parent friction coefficient ranges between 0 and 0.75, being the low the most common family of faults in the area, to study its influence values commonly used for developed fault zones and higher values on the neighboring segments. We also compute the stress change for less active faults (Deng and Sykes, 1997; Parsons et al., 1999; on optimally oriented faults at 2 km depth (Fig. 8) in order to study

ooo ooo ooo

o o

o o o o o o o o o

Fig. 8. Coulomb Failure Stress change at 2 km depth for optimally oriented A) strike slip faults, B) thrust faults and C) normal faults. The regional stress tensor is considered to have the main horizontal stress oriented NW–SE. The 2 km depth is selected in order to show the stress changes on the area where the main off-fault aftershock cluster took place. The circles show the aftershocks with hypocentral depths shallower than 3 km. The rectangle shows the projection of the main patch of the mainshock rupture. The surface traces of the structures are shown as thin lines. J.J. Martínez-Díaz et al. / Tectonophysics 546-547 (2012) 28–37 35 the distribution of a cluster of shallow off-fault aftershocks. The pos- 2010; Stein, 1999). The slip rate of the AMF has been calculated in itive values for ΔCFS are interpreted as promoting the faulting, previous works from neotectonic and paleoseismic records with a while negative values inhibit the activity. value ranging between 0.1 and 0.6 mm/year (Martinez-Diaz et al., About 75% of aftershocks accumulate mainly in the lower edge of 2010). The time of occurrence of an earthquake on a fault segment the rupture, presenting a good correlation with positive stress change undergoing tectonic loading is controlled both by the stress and fric- (Fig. 7). An exception to this rule is a cluster of aftershocks between tional properties on that fault and by earthquakes on other faults or 0 and 2 km deep (representing about 20% of the events). These after- segments nearby (Stein, 1999). For a fault loaded at a constant back- shocks appear to be associated with secondary fractures, not with the ground rate, a positive change of Coulomb failure stress produces a rupture fault plane, and maybe generated by a family of fractures time shift (advance) in the seismic cycle equivalent to: with different orientations. Models calculated on optimally oriented faults (Fig. 8) show that reverse faults oriented NE–SW receive the ΔT ¼ ΔCFS=τ; ð3Þ highest stress loading in the area where shallow off fault aftershocks accumulate. This family of reverse faults has been mapped in the zone where ΔT is the time shift, ΔCFS is the Coulomb failure stress change accompanied by folding (Figs. 2 and 9) and could be the origin of this and τ is the long term stressing rate on the fault. We model the long cluster of aftershocks. term stressing rate on the brittle part of the fault assuming a constant With respect to the influence of the Lorca earthquake in the stress displacement on the ductile deep segment of the fault (Stein et al., state of the surrounding segments of the AMF, our models indicate 1997; Toda et al., 1998). We obtain stress changes between 0.001 that the Goñar–Lorca segment is charged at the northern tip, and and 0.005 bar/year. A stress change of 1 bar represents a seismic the Lorca–Totana segment is charged at the southern tip. The stress cycle advance (time shift) equivalent to 200 to 1000 years of tectonic change on these segments exceeds 1 bar, a value that was shown to loading. be sufficient for the generation of earthquake triggering (Chen et al., 5. Discussion and conclusions A The proposed source model of the Lorca earthquake suggests that this event ruptured an area of ~4 by 3 km within a compressional strike slip duplex structure that limits the Goñar–Lorca and Lorca– Totana segments of the AMF (Figs. 2 and 9). The earthquake nucleat- ed to the north of this structure but most of the slip concentrates on a small area located to the SW of the hypocenter coinciding with the position of the duplex in the AMF (Fig. 9). This structure could act as an asperity during the rupture. Our results show a good agreement with the main, foreshock and aftershock relocations obtained by Lopez-Comino et al. (2012). The lateral position of maximum slip is coherent with the existence of an asymmetric bilateral rupture with 70% of the rupture propagating in SW direction described by these authors. Our Coulomb failure stress transfer models suggest that the 2011 Lorca earthquake induced static stress loading on the adjacent segments higher than 1 bar. This represents a seismic cycle advance equivalent to 200 to 1000 years of tectonic loading. This time shift could be significant in seismic hazard assessments and should be taken into account in future studies focused on this region. B Mapped tectonic structures in the epicentral region reveal the structural complexity of this intersegment zone, where the clear and well-defined trace of the Alhama de Murcia Fault in the adjacent segments loses continuity and is characterized by several structures with different dip and orientation. The junction of structures with different orientation results in kinematically complex areas (e.g. minor structures in Fig. 2) inducing regions of distributed deformation (process zone) (Deves et al., 2011; King, 1986; King and Nabelek, 1985). This complex structural configuration is unfavorable for the occurrence of large earthquakes, but favors the occurrence of small earthquakes, like the Lorca earthquake and its aftershocks. Interestingly, the topography near the fault shows a significant change near the intersegment zone (Fig. 2): both the Goñar–Lorca and Lorca–Totana segments present a significant positive relief to the NW (Las Estancias Range and La Tercia Range) associated with the inverse component of the strain accommodated in these faults. On the other hand the intersegment zone presents a depressed region to the NW, suggesting a Fig. 9. A) Interpretative block diagram, showing the position of the inferred fault rupture in relation to the AMF structure. Most of the Mo is released to the SW of the hypo- different way of long-term strain accommodation in this region. center (circle) coinciding with the position a complexity of the fault and the area of From our results, we propose a hypothesis for the long-term seis- higher deformation. B) Geological cross section through the maximum deformation mic activity of the Lorca intersegment zone according to which this area identified in the INSAR analysis. Thick line represents the estimated position of section of the AMF presents a characteristic seismic behavior driving the rupture plane. Thicker segment of this line indicates the area where most of the faulting and rupture styles that are different from those observed in slip concentrates (asperity). Crossed circle is the hypocenter of the Mainshock. Mate- – – rials: 1: basement rocks; 2: middle–upper Miocene to Pliocene deposits; 3: older Qua- the adjacent segments (Goñar Lorca and Lorca Totana). Deformation ternary deposits; 4: younger Quaternary (Late Pleistocene-Holocene) alluvial fans. is accommodated in these longer segments by localized slip in the 36 J.J. Martínez-Díaz et al. / Tectonophysics 546-547 (2012) 28–37 well-defined AMF and it builds the relief to the NW of the fault in the Deng, J., Sykes, L.R., 1997. Stress evolution in Southern California and triggering of moderate-, small-, and micro-sized earthquakes. Journal of Geophysical Research long-term. In the intersegment zone, deformation is accommo- 102, 411–435. dated in a distributed way, by slip of several structures with different DePolo, C.M., Clark, D.G., Slemmons, D.B., Aymand, W.H., 1989. Historical Basin and orientations (sources of small earthquakes, like Lorca and its after- Range Province surface faulting and fault segmentation. In: Schwartz, D.P., Sibson, R.H. (Eds.), Fault Segmentation and Controls of Rupture Initiation and Ter- shocks), preventing the building of relief. The Lorca intersegment re- mination: U. S. Geol. Surv. Open File Rep., 89–315, pp. 131–162. gion could act as a barrier for small‐moderate earthquakes (Mwb7), DePolo, C.M., Clark, D.G., Slemmons, D.B., Ramelli, A.R., 1991. Historical surface faulting but less frequent and larger earthquakes (Mw>7) would be capable in the Basin and Range Province, western North America—implications for fault – of propagating through the intersegment zone and rupturing both segmentation. Journal of Structural Geology 13, 123 136. Deves, M., King, G.C.P., Klinger, Y., Agnon, A., 2011. Localised and distributed deformation segments, as shown by available paleoseismic data (Masana et al., in the lithosphere: modelling the Dead Sea region in 3 dimensions. Earth and Plane- 2004; Ortuño et al., 2012). tary Science Letters 308, 172–184, http://dx.doi.org/10.1016/j.epsl.2011.05.044. fi In any case, the rupture process during the Lorca earthquake Doin, M.P., Lasserre, C., Peltzer, G., Doubre, C., 2009. Corrections of strati ed tropospheric delays in SAR interferometry: validation with global atmospheric models. Journal involved rupture directivity and heterogeneity in the slip distribu- of Applied Geophysics 69, 35–50, http://dx.doi.org/10.1016/j.jappgeo.2009.03.010. tion. High heterogeneity plays an important role in earthquake Du, Y., Aydin, A., Segall, P., 1992. Comparison of various inversion techniques as applied rupture propagation. Stored stress, strength, and frictional properties to the determination of a geophysical deformation model for the 1983 Borah Peak earthquake. Bulletin of the Seismological Society of America 82, 1840–1866. on the fault plane promote complexity in rupture processes (e.g. Page Dunham, E.M., Belanger, D., Cong, L., Kozdon, J.E., 2011. Earthquake ruptures with et al., 2005). The complex tectonic structure of the Lorca interseg- strongly rate-weakening friction and off-fault plasticity, part 2: nonplanar faults. ment area favors heterogeneous slip distributions and fluctuations Bulletin of the Seismological Society of America 101, 2308–2322. Elliott, J.R., Parsons, B., Jackson, J.A., Shan, X., Sloan, R.A., Walker, R.T., 2011. Depth segmen- in rupture velocity; both may contribute to damaging high frequency tation of the seismogenic continental crust: the 2008 and 2009 Qaidam earthquakes. ground motion (Boore and Joyner, 1978; Dunham et al., 2011; Geophysical Research Letters 38, L06305, http://dx.doi.org/10.1029/2011GL046897. Madariaga, 1977), as occurred in the Lorca earthquake. Elliott, J.R., Nissen, E.K., England, P.C., Jackson, J.A., Lamb, S., Li, Z., Oehlers, M., Parsons, fi B., 2012. Slip in the 2010–2011 Canterbury Earthquakes. New Zealand. Journal More paleoseismic data along the AMF, speci cally in the interseg- Geophysical Research, http://dx.doi.org/10.1029/2011JB008868. ment section of the fault, and tectonic geomorphology modeling are Fliss,S.,Bhat,H.S.,Dmowska,R.,Rice,J.R.,2005.Faultbranchingandrupturedirec- needed to investigate this hypothesis and to answer some arising tivity. Journal of Geophysical Research 110, B06312, http://dx.doi.org/10.1029/ questions: can we consider the Mw 5.2 Lorca earthquake as the 2004JB003368. Frontera, T., Concha, A., Blanco, P., Echeverria, A., Goula, X., Arbiol, R., Khazaradze, G., characteristic earthquake for this area? Which is the behavior of the Perez, F., Suriñach, E., 2012. DInSAR coseismic deformation of the May 2011 Mw intersegment region during large earthquakes that rupture both seg- 5.2 Lorca earthquake, (Southern Spain). Solid Earth Discussion 3, 111–119, ments? Can we expect peak slip at the intersegment zone as it has http://dx.doi.org/10.5194/se-3-111-2012. Gonzalez, P.J., Fernández, J., 2011. Drought-driven transient aquifer compaction im- been suggested for some other earthquakes (e.g. the 2008 Mw 7.9 aged using multitemporal satellite radar interferometry. Geology 39, 551–554, Wenchuan earthquake, Shen et al., 2009)? Or minimum local slip, as http://dx.doi.org/10.1130/G31900.1. suggested in other cases (e.g. the 2001 Mw 7.8 Kokoxili earthquake, Harris, R.A., 1998. Introduction to special section: stress triggers, stress shadows, and implications for seismic hazard. Journal of Geophysical Research 103 (B10), Klinger et al., 2005)? All these questions are of great importance for 24347–24358. seismic hazard assessment in the city of Lorca. Harris, R.A., Day, S.M., 1993. Dynamics of fault interaction: parallel strike-slip faults. Supplementary data related to this article can be found online at Journal of Geophysical Research 98, 4461–4472. Harris, R.A., Segall, P., 1987. Detection of a locked zone at depth on the Parkfield, Cali- http://dx.doi.org/10.1016/j.tecto.2012.04.010. fornia, segment of the San Andreas fault. Journal of Geophysical Research 92, 7945–7962. Acknowledgments Instituto Andaluz de Geofísica (IAG), 2011. Terremoto Lorca (11 Mayo 2011). Estudios preliminares, Granada, Spain. Available at http://www.ugr.es/iag. Instituto Geográfico Nacional (IGN), 2011. Serie terremoto NE Lorca (Murcia), Madrid. This research project was funded by Universidad Complutense de Available at http://www.ign.es. Madrid project TECTACT GR35/10-A-910368. We received financial sup- Instituto Geológico y Minero de España (IGME), 2011. Informe Geológico Preliminar del Terremoto de Lorca del 11 de Mayo del año 2011, 5.1 Mw, Madrid. Available port through Spanish projects CGL2008-01830, GEOTACTICA; CGL2009- at http://www.igme.es. 14405-C0202-BTE; Consolider CSD2006-00041 — TopoIberia and by the Jackson, J., Bouchon, M., Fielding, E., Funning, G., Ghorashi, M., Hatzfeld, D., Nazari, H., Junta de Andalucia project P09-RNM-5100, EC FP7 DORIS project (con- Parsons, B., Priestley, K., Talebian, M., Tatar, M., Walker, R., Wright, T., 2006. Seis- fi motectonic, rupture process, and earthquake-hazard aspects of the 2003 Decem- tract no. 242212) and nally we thank the European Space Agency ber 26 Bam, Iran, earthquake. Geophysical Journal International 166, 1270–1292. (ESA) for providing the ENVISAT images (Category1-9044 project). We King, G.C.P., 1986. Speculations on the geometry of the initiation and termination pro- utilized Coulomb 3 from the USGS to model Coulomb stress changes. cesses of earthquake rupture and its relation to morphology and geological structure. Pure and Applied Geophysics 124, 567–585. We thank Dr. Sylvain Barbot and Dr. Zheng-Kang Shen for their careful King, G.C., Nabelek, J., 1985. The role of fault bends in faults in the initiation and termi- revisions of the manuscript and for their contributions to improve this nation of earthquake rupture. Science 283, 984–987. work. Klinger, Y., 2010. Relation between continental strike-slip earthquake segmentation and thickness of the crust. Journal of Geophysical Research 115, B07306, http:// dx.doi.org/10.1029/2009JB006550. References Klinger, Y., Xiwei, Xu, Tapponnier, P., Van der Woerd, J., Lasserre, C., King, G., 2005. High-resolution satellite imagery mapping of the surface rupture and slip distribu- Aochi, H., Fukuyama, E., Matsumura, M., 2000. Spontaneous rupture propagation on a no- tion of the Mw 7.8, November 14, 2001, Kokoxili earthquake (Kunlun fault, north- planar fault in 3-D elastic medium. Pure and Applied Geophysics 157, 2003–2027. ern Tibet, China). Bulletin of the Seismological Society of America 95, 1970–1987. Beeler, N.M., Simpson, R.W., Hickman, S.H., Lockner, D.A., 2000. Pore fluid pressure, ap- Li, Z., Elliott, J.R., Feng, W., Jackson, J.A., Parsons, B., Walters, R., 2011. The 2010 Mw 6.8 parent friction, and Coulomb failure. Journal of Geophysical Research 105 (B11), Yushu (Qinghai, China) Earthquake: constraints provided by InSAR & Body Wave 25,533–25,542, http://dx.doi.org/10.1029/2000JB900119. Seismology. Journal of Geophysical Research 116, http://dx.doi.org/10.1029/ Boore, D.M., Joyner, W.B., 1978. The influence of rupture in coherence on seismic direc- 2011JB008358 16 pp. tivity. Bulletin of the Seismological Society of America 68, 283–300. Lohman, R.B., Simons, M., 2005. Some thoughts on the use of InSAR data to constrain Cavalie, O., Doin, M.P., Lasserre, C., Briole, P., 2007. Ground motion measurement in the models of surface deformation: noise structure and data down sampling. lake Mead area (Nevada, USA), by differential synthetic aperture radar interferom- Geochemistry, Geophysics, Geosystems 6, Q01007, http://dx.doi.org/10.1029/ etry time series analysis: probing the lithosphere rheological structure. Journal of 2004GC000841. Geophysical Research 112, B03403, http://dx.doi.org/10.1029/2006JB004344. Lopez-Comino, J.A., Mancilla, F., Morales, J., Stich, D., 2012. Rupture directivity of the Chen, K.H., Bürgmann, R., Nadeau, R.M., 2010. Triggering effect of M 4–5 earthquakes on 2011, Mw 5.2 Lorca earthquake (Spain). Geophysical Research Letters 39, the earthquake cycle of repeating events at Parkfield, California. Bulletin of the Seis- L03301, http://dx.doi.org/10.1029/2011GL050498. mological Society of America 100 (2), http://dx.doi.org/10.1785/0120080369. Lozos, J.C., Oglesby, D.D., Duan, B., Wenousky, S.G., 2011. The effects of double fault Cocco, M., Rice, J.R., 2002. Pore pressure and poroelasticity effects in Coulomb stress bends on rupture propagation: a geometrical parameter study. Bulletin of the Seis- analysis of earthquake interactions. Journal of Geophysical Research 107 (B2), mological Society of America 101, 385–398. 2030 AGU. Madariaga, R., 1977. High-frequency radiation from crack (stress drop) models of Das, S., Aki, K., 1977. A numerical study of two-dimensional spontaneous rupture prop- earthquake faulting. Geophysical Journal of the Royal Astronomical Society 51, agation. Geophysical Journal of the Royal Astronomical Society 50, 643–668. 625–651. J.J. Martínez-Díaz et al. / Tectonophysics 546-547 (2012) 28–37 37

Martinez-Diaz, J.J., 2002. Stress field variety related to fault interaction in a reverse the 2008 Wenchuan earthquake. Nature Geoscience 2, 718–724, http://dx.doi.org/ oblique-slip fault: the Alhama de Murcia Fault, Betic Cordillera, Spain. Tectonophy- 10.1038/ngeo636. sics 356, 291–305. Shengji, W., Fielding, E., Leprince, S., Sladen, A., Avouac, J.P., Helmberger, D., Hauksson, Martinez-Diaz, J.J., Masana, E., Hernández-Enrile, J.L., Santanach, P., 2001. Evidence for E., Chu, R., Simons, M., Hudnut, K., Herring, T., Briggs, R., 2011. Superficial simplicity coseismic events of recurrent prehistoric deformation along the Alhama de Murcia of the 2010 El Mayor-Cucapah earthquake of Baja California in Mexico. Nature fault, southeastern Spain. Geologica Acta 36, 315–327. Geoscience 4, 615–618. Martinez-Diaz, J.J., Masana, E., Ortuño, M., 2010. Implications of the structure of the Silva, P.G., Goy, J.L., Somoza, L., Zazo, C., Bardaji, T., 1993. Landscape response to strike- Alhama de Murcia Fault on its paleoseismological analysis. In: Insua, J.M., Martín- slip faulting linked to collisional settings: quaternary tectonics and basin formation González, F. (Eds.), Contribucion de la Geologia al Analisis de la Peligrosidad Sis- in the Eastern Betics, Southeastern Spain. Tectonophysics 224, 289–303. mica, Sigüenza (Guadalajara, Spain), pp. 97–100. Simons, M., Fialko, Y., Rivera, L., 2002. Coseismic deformation from the 1999 Mw 7.1 Masana, E., Martinez-Diaz, J.J., Santanach, P., Hernandez-Enrile, J.L., 2004. The Alhama Hector Mine, California, earthquake as inferred from InSAR and GPS observations. de Murcia Fault (SE Spain), a seismotectonic fault in a diffuse plate boundary. Seis- Bulletin of the Seismological Society of America 92, 1390–1402, http://dx.doi.org/ motectonic implications for the Iberomagrebian region. Journal of Geophysical Re- 10.1785/0120000933. search 109, B01301, http://dx.doi.org/10.1029/2002JB002359. Stein, R.S., 1999. The role of stress transfer in earthquake occurrence. Nature 402, Masana, E., Pallas, R., Perea, H., Ortuño, M., Martínez-Díaz, J.J., García-Meléndez, E., 605–609. Santanach, P., 2005. Large Holocene morphogenic earthquakes along the Albox Stein, R.S., Barka, A.A., Dieterich, J.H., 1997. Progressive failure on the North Anatolian fault, Betic Cordillera, Spain. Journal of Geodynamics 40, 119–133. fault since 1939 by earthquake stress triggering. Geophysical Journal International Nissen, E., Yamini-Fard, F., Tatar, M., Gholamzadeh, A., Bergman, E., Elliott, J.R., Jackson, 128, 594–604. J.A., Parsons, B., 2010. The vertical separation of mainshock rupture and microseis- Tarantola, A., Valette, B., 1982. Generalized nonlinear inverse problem solved using the micity at Qeshm Island in the Zagros fold-and-thrust belt, Iran. Earth and Planetary least squares criterion. Reviews of Geophysics 20, 219–232. Science Letters 296, 181–194, http://dx.doi.org/10.1016/j.epsl.2010.04.049. Toda, S., Stein, R.S., Reasenberg, P.A., Dieterich, J.H., Yoshida, A., 1998. Stress transferred Okada, Y., 1985. Surface deformation to shear and tensile faults in a half space. Bulletin by the 1995 Mw = 6.9 Kobe, Japan, shock: effect on aftershocks and future earth- of the Seismological Society of America 75, 1135–1154. quake probabilities. Journal of Geophysical Research 103 (B10), 24543–24565. Okada, Y., 1992. Internal deformation due to shear and tensile faults in a half-space. Vissers, R.L.M., Meijninger, B.M.L., 2011. The 11 May 2011 earthquake at Lorca (SE Bulletin of the Seismological Society of America 82, 1018–1040. Spain) viewed in a structural-tectonic context. Solid Earth Discussion 3, 527–540. Ortuño, M., Masana, E., García Meléndez, E., Martínez-Díaz, J., Štěpančíková, P., Cunha, Wesnousky, S., 2006. Predicting the end points of earthquake ruptures. Nature 444, P.P., Sohbati, R., Canora, C., Buylaert, J.P., Murria, A.S., 2012. An exceptionally long http://dx.doi.org/10.1038/nature05275. paleoseismic record of a slow-moving fault: the Alhama de Murcia fault (Eastern Zebker, H., Rosen, P., Goldstein, R.M., 1994. On the derivation of coseismic displace- Betic Shear Zone, Spain). Bulletin of the Geological Society of America. ment fields using differential radar interferometry: the Landers earthquake. Jour- doi:10.1130/B30558.1. nal of Geophysical Research 99, 19 617–19 634. Page, M.T., Dunham, E.M., Carlson, J.M., 2005. Distinguishing barriers and asperities in near-source ground motion. Journal of Geophysical Research 110, B11302, http:// dx.doi.org/10.1029/2005JB003736. Parsons, T., Stein, R.S., Simpson, R.W., Reasenberg, P.A., 1999. Stress sensitivity of fault REFERENCES CITED IN SUPPLEMENTARY DATA seismicity: a comparison between limited-offset oblique and major strike-slip faults. Journal of Geophysical Research 104 (B9), 20183–20202. Doin, M.P., Lasserre, C., Peltzer, G., Doubre, C., 2009b. Corrections of stratified tropospheric Reasemberg, P.A., Simpson, R.W., 1992. Response of regional seismicity to the static delays in SAR interferometry: validation with global atmospheric models. Journal of stress change produced by the Loma Prieta earthquake. Science 255, 1687–1690. Applied Geophysics 69, 35–50, http://dx.doi.org/10.1016/j.jappgeo.2009.03.010. Robinson, R., McGinty, P.J., 2000. The enigma of the Arthur's Pass, New Zealand, earth- Cavalie, O., Doin, M.P., Lasserre, C., Briole, P., 2007b. Ground motion measurement in quake 2. The aftershock distribution and its relation to regional and induced stress the lake Mead area (Nevada, USA), by differential synthetic aperture radar interfer- fields. Journal of Geophysical Research 105 (B7), 16139–16150. ometry time series analysis: probing the lithosphere rheological structure. Journal Scholz, C.H., 1990. Mechanics of Earthquakes and Faulting. Cambridge University Press. of Geophysical Research 112, B03403, http://dx.doi.org/10.1029/2006JB004344. 461pp. Zebker, H., Rosen, P., Goldstein, R.M., 1994b. On the derivation of coseismic displace- Shen, Z.-K., Sun, J., Zhang, P., Wan, Y., Wang, M., Bürgmann, R., Zeng, Y., Gan, W., Liao, ment fields using differential radar interferometry: the Landers earthquake. Jour- H., Wang, Q., 2009. Slip maxima at fault junctions and rupturing of barriers during nal of Geophysical Research 99, 19 617–19 634.